Dice games are attractive and many dice games are available or can be created easily. For a robot being able to play a dice game it means that it needs to be able to recognize, grasp, and throw a dice, in a robust manner in a natural indoor environment.
Training children with autism spectrum disorder using robots
Children with autism spectrum disorders (ASD) have marked impairments in social interaction and communication skills and have great difficulty in performing tasks that require initiation of and responding to social actions, such as imitation, turn taking, and collaborative (joint) action. Intensive behavioral interventions (1 therapist per child for 40 hours I week) in young children with ASD during one year or longer may substantially improve their social behavior and communication but are not cost-effective. The objective is to significantly improve the efficiency and quality in training social skills of children with ASD by developing personalized training scenarios for therapeutic robots in a natural home environment.
Collaborative research
The project on training children with ASD using robots is conducted in close collaboration with the technical university of Eindhoven, lead by Dr. Barakova and different clinics specialized in training children with ASD.
The role of TiViPE is to provide the robot infrastructure which includes the TiViPE graphical programming environment together with the latest software solutions on robotics.
Robot throwing dice
Children with ASD are intrigued by technology and humanoid robots catch their attention immediately. Many children love to play games and our aim is while playing games with a robot a social training scenario is used in such a way that a child with ASD is learning social skills while playing a game. A social training scenario is based upon a technique called Applied Behavior Analysis (ABA) and Pivotal Response Treatment (PRT) which is derived from ABA in particular is suitable for robot training scenario’s, since the steps taken follow a strict procedure.
Dice games are attractive and many dice games are available or can be created easily. For a robot being able to play a dice game it means that it needs to be able to recognize, grasp, and throw a dice, in a robust manner in a natural indoor environment.
Robot throwing dice in a technical perspective
TiViPE software combines, vision, together with its robotics textual language. In this approach
- A continuous stream of visual information is being processed.
- dice are being detected
- one every dice pips are recognized
- the position of the dice are transformed to real world coordinates
- A state space approach is being used.
- determination to grasp with left or right hand
- move arm up
- determination of the location on the table
- grasping
- holding dice evaluation
- throwing
- score evaluation
- user throwing dice
- user controlled evaluation
The application can be modified easily to perform automated evaluation, but also extended to play dice games.